Map subroutine implicit args to kernel's arg symbol, instead of copying (Attempt #2)

We don't need to copy the implicit arg since subroutines have access to the kernel's variables.
Just update the symbol map such that implicit arg intrinsics are mapped to the kernel's formal arguments.

Author: dlei6g

IGC/Compiler/CISACodeGen/EmitVISAPass.cpp

@@ -20844,16 +20844,9 @@ void EmitPass::emitImplicitArgIntrinsic(llvm::GenIntrinsicInst* I)
         IGC_ASSERT_MESSAGE(arg, "Implicit argument not found!");
         if (arg)
         {
-            if (I->getType()->isVectorTy())
-            {
-                emitVectorCopy(GetSymbol(I), m_currShader->getOrCreateArgumentSymbol(arg, false),
-                    int_cast<unsigned>(dyn_cast<IGCLLVM::FixedVectorType>(I->getType())->getNumElements()));
-            }
-            else
-            {
-                m_encoder->SetNoMask();
-                m_currShader->CopyVariable(GetSymbol(I), m_currShader->getOrCreateArgumentSymbol(arg, false));
-            }
+            // Direct symbol mapping to kernel argument, no need to copy for subroutines
+            CVariable* kernelArgV = m_currShader->getOrCreateArgumentSymbol(arg, false);
+            m_currShader->UpdateSymbolMap(I, kernelArgV);
         }
     }
     else